Mobile security

Can Mobile Networks Cope?

The summer of 2013 saw a number of headline-grabbing outages severely affecting the IT infrastructure of Internet giants such as Google, Amazon, Microsoft, and Apple, as well as Nasdaq. This has given rise to widespread soul-searching within and beyond the IT and telecom industries as to whether the pace unleashed by the digital and Internet networking revolutions may be proving too fast for the underlying network infrastructure to cope with. (See Outage Outrage.)

As part of that transformation, the mobile services ecosystem is undergoing its own profound changes. IP protocols are replacing legacy protocols at every layer from Layer 2 to Layer 7. And as the intermediary between the end-user and the Internet and applications environment, the mobile network is now in the eye of this storm of transformation, as evidenced by significant outages affecting T-Mobile US Inc. and Sprint Corp. (NYSE: S) during the summer. (See The Big Deal About Small Outages.)

Heavy Reading will be publishing a report of mine shortly on the subject of "Outages & Degradations In The Mobile Network." This is a fresh global survey of more than 70 qualified mobile operators investigating incidents of outages and degradations. How often do they occur? How long do they last? What causes them? How do operators deal with them? What impact do they have on the subscriber and on the operator?

I'm not about to go unloading all the key findings -- how well do you know us? -- but here are a couple to whet your appetite.

First, operators in developed markets identify an increase in the rate of subscriber churn as being the single biggest cost arising from incidents of outages and degradations in the mobile network.

Most operators are able to measure the impact of outages and degradations on subscriber churn. Indeed many see an impact immediately following a significant outage. This assertion may raise eyebrows among our older readers, but it's clear from the research that some age groups (particularly the 18-25s) are so dependent on their smartphones for the way that they live their lives that an hour or two's outage really can be enough to make them gnash their teeth, wail at the heavens, down smartphones, and switch operator.

Operators track churn, they correlate it to specific events such as outages and degradations, and they can demonstrate the correlation. They can also track the impact of outages and degradations on their Net Promoter Score (NPS), which is the extent to which consumers are willing (or unwilling) to recommend a consumer's product or service to their friends and family.

Another key finding is the very large gap between operators in developed markets and developing markets when it comes to outages and degradations. Operators in developing markets report substantially higher incident rates and a far greater concentration of incidents that take anywhere between one to four hours to put right. There are exceptions, of course. Some operators in developing markets -- for example in the Middle East -- have invested very heavily not just in state-of-the-art infrastructure throughout the network, but also in the area of leading edge network monitoring systems which can assist in alerting the operator to actual and impending incidents as well as in remediation and root cause analysis.

If you take an average, then it would appear that mobile operators are actually coping fairly well in terms of network uptime and network performance in the mobile broadband era. But that average conceals very significant differences between the operators. There are clearly some that are managing to keep outages and degradations down to a minimum. Then there are other others that, on a daily basis, are struggling -- and in many cases failing -- to keep the network up and the churn rate down.

— Patrick Donegan, Senior Analyst, Heavy Reading

jhodgesk1s 10/15/2013 | 11:18:58 AM
Re: Diameter? Patrick, agreed, and clearly the move to a dedicated Diameter Signaling Controller (DSC) architecture is part of the solution.
DanJones 10/15/2013 | 10:31:35 AM
Re: 'How do operators deal with them?' That's never a fast process though right? They were talking standards what 2-3 years ago.
DanJones 10/15/2013 | 10:29:27 AM
Re: 'How do operators deal with them?' I wouldn't say that's unique to AT&T. Say as little as possible seems to the operator mantra for outages.
pdonegan67 10/15/2013 | 10:27:36 AM
Re: 'How do operators deal with them?' A core part of the survey report, yes. Report's out soon. I'm sure we can cut you a 10% discount, Ray. Maybe even 20%.
[email protected] 10/15/2013 | 10:25:24 AM
'How do operators deal with them?' "How do operators deal with them?"

That's at the center of all this, right?

I hope they can evolve their strategies around this - I seem to recall that AT&T dealt with its most recent major outage by sticking its head in the sand and ignoring hte social networking storm....

Looking forward to hearing more on this. 
pdonegan67 10/15/2013 | 10:24:40 AM
Re: Diameter? One among many causes, yes, but there's a lot of work going on among vendors and in the standards bodies to enable Diameter to cope better in high loading situations so that will certainly help where that's concerned.
DanJones 10/15/2013 | 10:22:08 AM
Diameter? How many LTE outages are down to the sheer number of Diameter calls on the systems?

Its been suggested to me that Diameter calls are one of the major causes of woe in new 4G networks by several different vendors. True?
Sign In